Credit risk modelling is an approach to identify high risk customers. Credit scoring is based on analysing customer’s personal information, transactional information, last premiums and many more variables providing a unique score reflecting the creditworthiness of the customer.
With credit scoring, TransOrg helps in building customer application evaluation, collection strategies, cross/up-sell models, customer acquisition and retention strategies.
Data stitching from various sources
Understand the data schema and data dictionary.
Access data quality gaps and data sufficiency.
Perform vintage and roll-rate analysis, if needed.
Perform variable transformation and correlation tests for multicollinearity.
Feature selection based on variable significance and intelligence value
Model methodology specification and implementation
Build machine learning risk scoring classifier model
Carry out scorecard validation and out of time validation
Perform reject inferencing, if needed
Review Model approach
Access model implementation
